Learning produces long-term changes in behavior, physiology, and gene expression. Although this principle is now well-documented across the animal kingdom, fundamental aspects of learning and memory remain unclear, including: 1) how the transcriptional changes that maintain long-term memory are sustained and 2) if forgetting is an active process or merely the passive decay of maintenance processes. We will address these issues through the study of long-term sensitization (LTS) in the marine gastropod Aplysia californica. Animals will receive long-term sensitization training, which produces a long-lasting but unilateral memory expressed as an increase in the duration of the tail-elicited siphon-withdrawal reflex (T-SWR). Changes in gene expression will then be measured via microarray when the memory is strongly expressed (1 day after training, Aim 1) or forgotten (1 week after training, Aim 2). How are memories maintained? (Aim 1) Identification of transcriptional changes that persist 1 day after training will provide insight into the maintenance of long-term memory. We will use this data to develop a model of memory maintenance. Specifically, we will utilize a bioinformatics approach (FIRE) to discover promoter motifs specific to LTS memory maintenance. qPCR will be used to confirm regulation of the functional clusters identified by microarray and bioinformatics. We will then use a neural analogue of sensitization to probe the functional significance of these learning-regulated motifs. Are memories erased or do they just fade away? (Aim 2) Comparison of transcriptional 7 days after training will reveal if forgetting is associated with the simple decline of maintenance-related changes (as predicted by passive decay models of forgetting) or with the induction of additional transcriptional changes (as predicted by active forgetting models). Our microarray findings will provide data in support of one of these two possible mechanisms. If passive decay is supported by our data, we will explore the time-course of transcriptional decay by comparing expression changes via qPCR at 3, 5 and 7 days after LTS training. If active forgetting is supported, we will use the same time-course to delineate the activation of forgetting-related transcripts. This project is based on strong pilot data and well-validated methodologies; it provides a tractable means to gain insight into the mechanisms of long-term memory in a straightforward experimental context that offers excellent opportunities for undergraduate involvement.